Endometriosis Apps, AI Help Researchers Understand the Condition

Noémie Elhadad, PhD, associate professor and chair of biomedical informatics at Columbia University’s Vagelos College of Physicians and Surgeons, knows all too well the limited information available about endometriosis. She was diagnosed with the condition at the age of 13—much earlier than most women who are diagnosed in their 20s or 30s. 

Her career has been dedicated to learning more about the chronic condition that affects one in 10 women of child-bearing age plus trans and nonbinary people with uteruses. Among her contributions is a research mobile app that has 17,000 users who submit information about their experiences with endometriosis.  

Noemie Elhadad

Noémie Elhadad. Photo by Jörg Meyer.

Elhadad developed the app, Phendo, to gather data from patients around the world who suffer from endometriosis. “The primary goal is to have a registry of patients and their day-to-day symptoms, the types of treatments they’re using—supplements, exercise, food that triggers symptoms—anything they can tell us about the disease and how their experience of disease changes from one day to the next,” says Elhadad.  

People often experience debilitating pain for years before seeking help, and not all patients with endometriosis have pain. “Because pain can seem so subjective, and there is still a lot of open questions about the etiology of endometriosis, endometriosis is not always the first thing that comes to mind for patients or their doctors.”  

Because symptoms vary widely, patients can experience years-long delays in diagnosis with the average diagnosis coming up to 10 years after the onset of symptoms. 

Endometriosis is not a disease of painful menstruation alone. In many patients, endometriosis causes gastrointestinal and genitourinary problems as well as chronic pain in many areas of the body. “The data patients contributed to us are confirming the suspicions of a lot of researchers that endometriosis has an inflammatory component,” says Elhadad.  

In addition to trying to create better descriptions of the disease itself, Elhadad and her team want to build tools to help patients manage their symptoms. “Management of endometriosis is complex. Patients and their care team are not always aligned, and guidelines for endometriosis management are not specific because there is so much variation between what works from one patient to another.” By synthesizing the patients’ day-to-day experiences of diseases, the app can help patients facilitate discussions with their care team and ground the discussions in these experiences.   

But learning what strategies work for one person does not translate into recommendations for others with similar characteristics. “We don’t know yet what it means for a patient to be similar to another patient,” Elhadad says. “For instance, physical exercise helps some patients tremendously with their pelvic pain. For others, it exacerbates their symptoms.”  

To develop more personalized self-management strategies, Elhadad’s lab uses an AI-based technology called reinforcement learning. “It learns what happens to a patient as a result of different actions and determines which are the best actions to bring the patient closer to a goal such as reduction of pain or fatigue,” she says.  

Elhadad is also developing an AI-based tool that could lead to improved screening for the condition. The goal is to review electronic health records to seek a “signature” that could identify potential cases. Rather than diagnosing patients automatically, the goal of this AI is to help patients and their providers connect the dots and recommend a conversation with specialty gynecology about endometriosis.